Hough transform – detecting lines and circles

In image processing, Hough transform is a feature extraction technique that aims to find instances of objects of a certain shape using a voting procedure carried out in a parameter space. In its simplest form, the classical Hough transform can be used to detect straight lines in an image. We can represent a straight line using polar parameters (ρ, θ), where ρ is the length of the line segment and θ is the angle in between the line and the x axis. To explore (ρ, θ) parameter space, it first creates a 2D-histogram. Then, for each value of ρ and θ, it computes the number of non-zero pixels in the input image that are close to the corresponding line and increments the array at position (ρ, θ) accordingly. Hence, each non-zero pixel can be thought of as voting for potential line candidates. The most probable lines correspond to the parameter values that obtained the highest votes, that is, the local maxima in a 2D histogram. The method can be extended...